The problem inherent in google is that it will not distinguish between posts on the same page. It will tend to rank based on keyword proximity, which might give posts matching all keywords some ranking advantage, but it will not completely throw out results where, say, .38 and super occur on the same page of a thread but in different posts.
If you want to search for a phrase, like ".38 super", -then- you can use google to get reliable results. Since you're searching for one phrase and not multiple keywords, google's inability to tell posts apart won't matter.
The search engine TFL uses is the best I've found so far. It has its limitations in features, but when the alternatives are a) use google; or b) throw more servers at the problem, I think this does pretty well. It performs extremely well.
Don't forget there are also privacy implications from outsourcing searching to google. Some people hate google with a passion. What would they do if TFL's search form went direct to google?
Quote:
|
However, the software is pulling up anything that 38 OR super so it pulls up a ton of irrelevant results.
|
Are you sure? I did a quick spot check and search results for .38 super seem to include both words. The engine doesn't support phrase/proximity searching, so ".38 super" won't work like it would with mysql fulltext, but in that case you can use google.